Code for physics-closed-loop manifold transfer of square probes (COMSOL+MATLAB+Python)
Abstract
This record contains the MATLAB and Python scripts used for the physics-closed-loop active learning workflow in this study. The Python script trains a physics-informed VAE and generates candidate designs, while the MATLAB script evaluates candidates in COMSOL via mphserver (LiveLink for MATLAB). Requirements: Python (numpy, pandas, torch, scikit-learn), MATLAB, and COMSOL Multiphysics with LiveLink for MATLAB.Note: COMSOL/Matlab paths in the scripts are configured for macOS and may need to be adapted.
Fields of Study & Subjects
Model Reduction and Neural Networks
Machine Learning in Materials Science
Advanced Multi-Objective Optimization Algorithms
Publication details
- Journal / source
- Zenodo (CERN European Organization for Nuclear Research)
- Publication date
- 3 February 2026